SARSA.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np \n",
    "import gym\n",
    "from gym import spaces\n",
    "import random as rd\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "from envs.NetworkOperatorEnv import NetworkOperatorEnv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "\n",
    "def eps_greedy(Q, s, eps=0.1):\n",
    "    '''\n",
    "    Epsilon greedy policy\n",
    "    '''\n",
    "    if np.random.uniform(0,1) < eps:\n",
    "        # Choose a random action\n",
    "        return np.random.randint(Q.shape[1])\n",
    "    else:\n",
    "        # Choose the action of a greedy policy\n",
    "        return greedy(Q, s)\n",
    "\n",
    "\n",
    "def greedy(Q, s):\n",
    "    '''\n",
    "    Greedy policy\n",
    "    return the index corresponding to the maximum action-state value\n",
    "    '''\n",
    "    return np.argmax(Q[s])\n",
    "\n",
    "\n",
    "def run_episodes(env, Q, num_episodes=100, to_print=False):\n",
    "    '''\n",
    "    Run some episodes to test the policy\n",
    "    '''\n",
    "    tot_rew = []\n",
    "    state = env.reset()\n",
    "\n",
    "    for _ in range(num_episodes):\n",
    "        done = False\n",
    "        game_rew = 0\n",
    "\n",
    "        while not done:\n",
    "            # select a greedy action\n",
    "            next_state, rew, done, _ = env.step(greedy(Q, state))\n",
    "\n",
    "            state = next_state\n",
    "            game_rew += rew \n",
    "            if done:\n",
    "                state = env.reset()\n",
    "                tot_rew.append(game_rew)\n",
    "\n",
    "    if to_print:\n",
    "        print('Mean score: %.3f of %i games!'%(np.mean(tot_rew), num_episodes))\n",
    "\n",
    "    return np.mean(tot_rew)\n",
    "\n",
    "\n",
    "def SARSA(env, lr=0.01, num_episodes=10000, eps=0.3, gamma=0.95, eps_decay=0.00005):\n",
    "    nA = env.action_space.n\n",
    "    nS = env.observation_space.n\n",
    "    \n",
    "    \n",
    "    ind_ep =  []\n",
    "    costs =  []\n",
    "\n",
    "    # Initialize the Q matrix\n",
    "    # Q: matrix nS*nA where each row represent a state and each colums represent a different action\n",
    "    Q = np.zeros((nS, nA))\n",
    "    rewards = []\n",
    "\n",
    "    for ep in range(num_episodes):\n",
    "        state = env.reset()\n",
    "        done = False\n",
    "        tot_rew = 0\n",
    "\n",
    "        # decay the epsilon value until it reaches the threshold of 0.01\n",
    "        if eps > 0.01:\n",
    "            eps -= eps_decay\n",
    "\n",
    "        action = eps_greedy(Q, state, eps) \n",
    "        \n",
    "        next_state, rew, done, _ = env.step(action) # Take one step in the environment\n",
    "    \n",
    "        # choose the next action (needed for the SARSA update)\n",
    "        next_action = eps_greedy(Q, next_state, eps) \n",
    "        \n",
    "        # SARSA update\n",
    "        Q[state][action] = Q[state][action] + lr*(rew + gamma*Q[next_state][next_action] - Q[state][action])\n",
    "\n",
    "        state = next_state\n",
    "        action = next_action\n",
    "        tot_rew += rew\n",
    "        #if done:\n",
    "        rewards.append(tot_rew)\n",
    "        \n",
    "        if (ep % 200) == 0:\n",
    "            print(\"Episode:{:5d}  Eps:{:2.4f}  Rew:{:2.4f}\".format(ep, eps, tot_rew))\n",
    "            ind_ep.append(ep)\n",
    "            costs.append(tot_rew)\n",
    "    \n",
    "    print(Q)\n",
    "    plt.plot(ind_ep, costs)\n",
    "    #plt.xlim(4000, 5000)\n",
    "    plt.title('SARSA Algorithm for Resource Allocation' )\n",
    "    plt.xlabel('Number of episodes')\n",
    "    plt.ylabel('Cost')\n",
    "    plt.grid('on')\n",
    "    #plt.rcParams[\"figure.figsize\"] = [16, 9]\n",
    "    plt.show()  \n",
    "    \n",
    "    \n",
    "        # Test the policy every 300 episodes and print the results\n",
    "   #     if (ep % 300) == 0:\n",
    "   #         test_rew = run_episodes(env, Q, 1000)\n",
    "   #         print(\"Episode:{:5d}  Eps:{:2.4f}  Rew:{:2.4f}\".format(ep, eps, test_rew))\n",
    "   #         test_rewards.append(test_rew)\n",
    "\n",
    "    #return Q"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Episode:    0  Eps:0.4990  Rew:10.0000\n",
      "Episode:  200  Eps:0.2990  Rew:41.0000\n",
      "Episode:  400  Eps:0.0990  Rew:44.0000\n",
      "Episode:  600  Eps:0.0100  Rew:37.0000\n",
      "Episode:  800  Eps:0.0100  Rew:42.0000\n",
      "Episode: 1000  Eps:0.0100  Rew:6.0000\n",
      "Episode: 1200  Eps:0.0100  Rew:37.0000\n",
      "Episode: 1400  Eps:0.0100  Rew:37.0000\n",
      "Episode: 1600  Eps:0.0100  Rew:40.0000\n",
      "Episode: 1800  Eps:0.0100  Rew:40.0000\n",
      "Episode: 2000  Eps:0.0100  Rew:49.0000\n",
      "Episode: 2200  Eps:0.0100  Rew:49.0000\n",
      "Episode: 2400  Eps:0.0100  Rew:40.0000\n",
      "Episode: 2600  Eps:0.0100  Rew:24.0000\n",
      "Episode: 2800  Eps:0.0100  Rew:29.0000\n",
      "Episode: 3000  Eps:0.0100  Rew:16.0000\n",
      "Episode: 3200  Eps:0.0100  Rew:22.0000\n",
      "Episode: 3400  Eps:0.0100  Rew:34.0000\n",
      "Episode: 3600  Eps:0.0100  Rew:22.0000\n",
      "Episode: 3800  Eps:0.0100  Rew:33.0000\n",
      "Episode: 4000  Eps:0.0100  Rew:37.0000\n",
      "Episode: 4200  Eps:0.0100  Rew:45.0000\n",
      "Episode: 4400  Eps:0.0100  Rew:36.0000\n",
      "Episode: 4600  Eps:0.0100  Rew:40.0000\n",
      "Episode: 4800  Eps:0.0100  Rew:38.0000\n",
      "Episode: 5000  Eps:0.0100  Rew:35.0000\n",
      "Episode: 5200  Eps:0.0100  Rew:30.0000\n",
      "Episode: 5400  Eps:0.0100  Rew:24.0000\n",
      "Episode: 5600  Eps:0.0100  Rew:36.0000\n",
      "Episode: 5800  Eps:0.0100  Rew:29.0000\n",
      "Episode: 6000  Eps:0.0100  Rew:35.0000\n",
      "Episode: 6200  Eps:0.0100  Rew:39.0000\n",
      "Episode: 6400  Eps:0.0100  Rew:35.0000\n",
      "Episode: 6600  Eps:0.0100  Rew:34.0000\n",
      "Episode: 6800  Eps:0.0100  Rew:42.0000\n",
      "Episode: 7000  Eps:0.0100  Rew:40.0000\n",
      "Episode: 7200  Eps:0.0100  Rew:22.0000\n",
      "Episode: 7400  Eps:0.0100  Rew:47.0000\n",
      "Episode: 7600  Eps:0.0100  Rew:35.0000\n",
      "Episode: 7800  Eps:0.0100  Rew:32.0000\n",
      "Episode: 8000  Eps:0.0100  Rew:41.0000\n",
      "Episode: 8200  Eps:0.0100  Rew:35.0000\n",
      "Episode: 8400  Eps:0.0100  Rew:49.0000\n",
      "Episode: 8600  Eps:0.0100  Rew:29.0000\n",
      "Episode: 8800  Eps:0.0100  Rew:49.0000\n",
      "Episode: 9000  Eps:0.0100  Rew:32.0000\n",
      "Episode: 9200  Eps:0.0100  Rew:37.0000\n",
      "Episode: 9400  Eps:0.0100  Rew:30.0000\n",
      "Episode: 9600  Eps:0.0100  Rew:31.0000\n",
      "Episode: 9800  Eps:0.0100  Rew:36.0000\n",
      "Episode:10000  Eps:0.0100  Rew:38.0000\n",
      "Episode:10200  Eps:0.0100  Rew:40.0000\n",
      "Episode:10400  Eps:0.0100  Rew:25.0000\n",
      "Episode:10600  Eps:0.0100  Rew:29.0000\n",
      "Episode:10800  Eps:0.0100  Rew:22.0000\n",
      "Episode:11000  Eps:0.0100  Rew:24.0000\n",
      "Episode:11200  Eps:0.0100  Rew:20.0000\n",
      "Episode:11400  Eps:0.0100  Rew:48.0000\n",
      "Episode:11600  Eps:0.0100  Rew:22.0000\n",
      "Episode:11800  Eps:0.0100  Rew:32.0000\n",
      "Episode:12000  Eps:0.0100  Rew:41.0000\n",
      "Episode:12200  Eps:0.0100  Rew:13.0000\n",
      "Episode:12400  Eps:0.0100  Rew:45.0000\n",
      "Episode:12600  Eps:0.0100  Rew:45.0000\n",
      "Episode:12800  Eps:0.0100  Rew:26.0000\n",
      "Episode:13000  Eps:0.0100  Rew:24.0000\n",
      "Episode:13200  Eps:0.0100  Rew:30.0000\n",
      "Episode:13400  Eps:0.0100  Rew:36.0000\n",
      "Episode:13600  Eps:0.0100  Rew:8.0000\n",
      "Episode:13800  Eps:0.0100  Rew:33.0000\n",
      "Episode:14000  Eps:0.0100  Rew:33.0000\n",
      "Episode:14200  Eps:0.0100  Rew:40.0000\n",
      "Episode:14400  Eps:0.0100  Rew:27.0000\n",
      "Episode:14600  Eps:0.0100  Rew:31.0000\n",
      "Episode:14800  Eps:0.0100  Rew:35.0000\n",
      "Episode:15000  Eps:0.0100  Rew:40.0000\n",
      "Episode:15200  Eps:0.0100  Rew:51.0000\n",
      "Episode:15400  Eps:0.0100  Rew:54.0000\n",
      "Episode:15600  Eps:0.0100  Rew:37.0000\n",
      "Episode:15800  Eps:0.0100  Rew:40.0000\n",
      "Episode:16000  Eps:0.0100  Rew:40.0000\n",
      "Episode:16200  Eps:0.0100  Rew:27.0000\n",
      "Episode:16400  Eps:0.0100  Rew:33.0000\n",
      "Episode:16600  Eps:0.0100  Rew:43.0000\n",
      "Episode:16800  Eps:0.0100  Rew:34.0000\n",
      "Episode:17000  Eps:0.0100  Rew:35.0000\n",
      "Episode:17200  Eps:0.0100  Rew:26.0000\n",
      "Episode:17400  Eps:0.0100  Rew:18.0000\n",
      "Episode:17600  Eps:0.0100  Rew:37.0000\n",
      "Episode:17800  Eps:0.0100  Rew:9.0000\n",
      "Episode:18000  Eps:0.0100  Rew:23.0000\n",
      "Episode:18200  Eps:0.0100  Rew:38.0000\n",
      "Episode:18400  Eps:0.0100  Rew:33.0000\n",
      "Episode:18600  Eps:0.0100  Rew:32.0000\n",
      "Episode:18800  Eps:0.0100  Rew:32.0000\n",
      "Episode:19000  Eps:0.0100  Rew:29.0000\n",
      "Episode:19200  Eps:0.0100  Rew:49.0000\n",
      "Episode:19400  Eps:0.0100  Rew:28.0000\n",
      "Episode:19600  Eps:0.0100  Rew:22.0000\n",
      "Episode:19800  Eps:0.0100  Rew:37.0000\n",
      "[[414.62260333   0.           0.         ... 508.68571422 581.49551664\n",
      "    0.        ]\n",
      " [  0.         562.49162682   0.         ...   0.           0.\n",
      "   57.03893364]\n",
      " [442.37441188 546.85871173   0.         ...   0.           0.\n",
      "    0.        ]\n",
      " ...\n",
      " [505.41054534   0.           0.         ...   0.           0.\n",
      "    0.        ]\n",
      " [  0.           0.           0.         ...   0.           0.\n",
      "    0.        ]\n",
      " [545.09996274   0.           0.         ...   0.           0.\n",
      "    0.        ]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "env = NetworkOperatorEnv(3, 30)\n",
    "SARSA(env, lr=.9, num_episodes=20000, eps=0.5, gamma=0.95, eps_decay=0.001)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}